This is a repository copy of Switching Aurora-A kinase on and off at an allosteric site. White Rose Research Online URL for this paper: http://eprints.whiterose.ac.uk/115219/ Version: Accepted Version Article: Bayliss, R orcid.org/0000-0003-0604-2773, Burgess, SG and McIntyre, PJ (2017) Switching Aurora-A kinase on and off at an allosteric site. The FEBS Journal, 284 (18). pp. 2947-2954. ISSN 1742-464X https://doi.org/10.1111/febs.14069 (c) 2017, Federation of European Biochemical Societies. This is the peer reviewed version of the following article: 'Bayliss, R , Burgess, SG and McIntyre, PJ (2017) Switching Aurora-A kinase on and off at an allosteric site. The FEBS Journal, 284 (18). pp. 2947-2954,' which has been published in final form at [https://doi.org/10.1111/febs.14069]. This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Self-Archiving. Reuse Unless indicated otherwise, fulltext items are protected by copyright with all rights reserved. The copyright exception in section 29 of the Copyright, Designs and Patents Act 1988 allows the making of a single copy solely for the purpose of non-commercial research or private study within the limits of fair dealing. The publisher or other rights-holder may allow further reproduction and re-use of this version - refer to the White Rose Research Online record for this item. Where records identify the publisher as the copyright holder, users can verify any specific terms of use on the publisher’s website. Takedown If you consider content in White Rose Research Online to be in breach of UK law, please notify us by emailing [email protected] including the URL of the record and the reason for the withdrawal request. [email protected] https://eprints.whiterose.ac.uk/ Switching Aurora-A kinase on and off at an allosteric site Richard Bayliss1*, Selena G. Burgess1, Patrick J. McIntyre2 1. Astbury Centre for Structural Molecular Biology, Faculty of Biological Sciences, University of Leeds, Leeds, UK. 2. Department of Molecular and Cell Biology, University of Leicester, Leicester, UK. Corresponding Author: Richard Bayliss Email: [email protected] Phone: (44) 113 3439919 Keywords: Allostery, protein-protein interaction, protein kinase, kinase inhiBitor Abbreviations: PKA – protein kinase A, ATP – adenosine triphosphate, PPI – protein- protein interaction, AGC – protein kinase A, G, C family, sdAB – single domain antiBody, HF – hydrophoBic motif, vNAR – variable domain new antigen receptor, PDB – protein data Bank, HV2 – hypervariable loop 2, CDR1 - complementarity-determining region 1, PP6 - protein phosphatase 6, PP1 – protein phosphatase 1, TPX2 – targeting protein for Xklp2, TACC3 – transforming acidic coiled-coil containing protein 3, PDK1 – phosphoinositide-dependent kinase-1, PKB/Akt – protein kinase B, RSK – riBosomal s6 kinase, CDK – cyclin-dependent kinase, EGFR – epidermal growth factor receptor, ADP – adenosine diphosphate. Abstract Protein kinases are central players in the regulation of cell cycle and signalling pathways. Their catalytic activities are strictly regulated through post-translational modifications and protein-protein interactions that control switching Between inactive and active states. These states have Been studied extensively using protein crystallography, although the dynamic nature of protein kinases makes it difficult to capture all relevant states. Here we descriBe two recent structures of Aurora-A kinase that trap its active and inactive states. In Both cases, Aurora-A is trapped through interaction with a synthetic protein, either a single domain antiBody that inhiBits the kinase or a hydrocarBon-stapled peptide that activates the kinase. These structures show how the distinct synthetic proteins target the same allosteric pocket with opposing effects on activity. These studies pave the way for the development of tools to proBe these allosteric mechanisms in cells. Introduction Many of the key signalling and regulatory events inside cells involve protein phosphorylation, a reversiBle and uBiquitous post-translational modification carried out By protein kinases. The human genome encodes more than 500 protein kinase domains, most of which are found in the context of a multidomain polypeptide chain, and many are components of multi-protein complexes [1]. The catalytic activities of protein kinases are regulated through interactions with these other domains and also By post-translational modifications [2]. Protein-protein interactions also serve to localise signalling pathways to specific suB-cellular localisations to defined multi-protein complexes that assemBle at the sites of kinase activity. Protein kinases have conserved sequence and structural features that are organised into a similar conformational arrangement when the kinase is in an active state (Figure 1a,B). In contrast, these features are displaced in many different ways when a kinase is inactive. The catalytic domain of a protein kinase comprises two suB-domains or loBes, Between which a molecule of ATP Binds. The N-loBe is formed from a b-sheet and at least one a-helix, usually called the C-helix Based on the canonical PKA structure [3]. The C-loBe is mostly helical and has an extended loop structure called the activation loop, activation segment or T-loop, which is the site of regulatory phosphorylation and forms part of the Binding site for protein suBstrates. The structural Basis of protein kinase activation has Been the suBject of many research articles, and has Been extensively reviewed elsewhere [2, 4-7]. Therefore, for reasons of Brevity we will focus on a suBset of concepts relevant to the allosteric regulation of Aurora-A. Three features in particular are hallmarks of kinase active or inactive states: a Lys-Glu salt-Bridge, the regulatory (R-) spine and the activation loop (Figure 1a,B). The Lys- Glu salt Bridge Brings together the b-sheet and C-helix and helps to position the phosphates of ATP for catalysis [2, 5]. The R-spine is a set of four hydrophoBic residues that connect these same two major structural elements of the N-loBe with the C-loBe. The side chains of the R-spine residues form a connected, roughly linear pattern in active kinases, But are displaced from this arrangement in inactive kinases [4, 6]. Structural and Biochemical studies have defined the molecular Basis of protein-protein interactions (PPIs) involving kinases. As well as kinase-suBstrate interactions, regulatory interactions of kinases may involve symmetric or asymmetric dimerization or interactions with other Binding partners. Studies of protein-protein interactions typically rely on genetic approaches. Small molecule inhiBitors offer greater flexiBility in terms of application to different cell types and the ability to deliver a known dose at a specified time point. However, the development of small molecule PPI inhiBitors has considerable time and monetary costs. Moreover, the expertise required to carry out chemical synthesis is not usually available in a typical molecular, cell or structural Biology laboratory. We therefore Began to exploit the potential of peptide and protein-Based reagents to target the protein- protein interactions of a model Ser/Thr protein kinase, Aurora-A, a key regulator of cell division [8]. Phosphorylation and interactions coordinate to activate Aurora-A Aurora-A is related to the AGC family of Ser/Thr kinases, which are activated By phosphorylation of a site in their activation loop and, in many cases, one or more sites in a region C-terminal to the catalytic domain [9]. A hallmark of AGC kinases is their activation By a hydrophoBic (HF) motif within a C-terminal extension to the kinase domain that fits into a hydrophoBic surface on the N-loBe Between the b-sheet and the C-helix called the PIF pocket (Figure 1c) [9, 10]. In addition, some AGC kinases have a helical N-terminal extension from which Trp and Phe (F/W) side chains protrude into a groove Between the C-helix and activation loop (Figure 1c) [11]. PDK1 is an AGC kinase that lacks a HF motif and, instead, its PIF pocket is used to recruit other AGC kinases, such as PKB/AKT and RSK, which are then phosphorylated and activated [9]. This property has Been exploited in the development of small molecules that Bind to the PIF pocket and Block interactions and, interestingly, this same pocket can Be used to activate or inhiBit PDK1 [12-14]. Aurora-A has a similar core catalytic domain to AGC kinases, But lacks the N- and C-terminal extensions (Figure 1d). Instead, the HF motif and F/W side chains are provided By a Binding partner, the microtuBule-associated protein TPX2 (Figure 1e,f) [15]. HF motifs and the region of TPX2 that interact with the hydrophoBic pocket are short extended peptides. However, the same surface area of the catalytic domain of many other kinases is involved in regulatory interactions with an a-helix on the surface of a folded partner protein, for example the Binding of cyclins to CDKs, and the asymmetric dimerization of EGFR. The hydrophoBic pocket of kinases is therefore a versatile Binding site, able to accommodate an extended peptide, an a-helix or an organic compound of low molecular weight. Aurora-A has an intrinsically low level of catalytic activity that is Boosted through autophosphorylation on Thr288, one of a pair of adjacent threonine residues in a region of the kinase called the activation loop [16]. The mechanism of autophosphorylation remains undecided, with Biochemical evidence in support of Both intramolecular and intermolecular mechanisms [17, 18]. Aurora-A activity is also promoted through interaction with protein Binding partners, such as the microtuBule-associated proteins TPX2 and TACC3, the centrosomal protein Cep192, the ciliary protein Pifo and the transcription factor N-Myc [15, 19-23]. These regulatory partners stimulate autophosphorylation and, at least in the case of TPX2 in vitro, these two activation steps can comBine synergistically to produce a hyperactive form of the kinase (Figure 2a) [24].